Stewart D. Cusworth

Misalignment losses at multimode graded-index fiber splices and GRIN rod lens couplers

A geometric optics model to describe transverse offset splice loss in multimode graded-index fiber, which has been adapted to evaluate misalignment losses at a graded-index rod (GRIN rod) lens-to-lens coupler, is reported. In both cases, the model gives close agreement with experimental measurements. Finally, in contrast with fiber splice losses, invariance of the lens coupler loss with the optical launch and receive conditions is shown.

Angular tilt misalignment loss at a GRIN rod lens coupler.

A geometric optics model for evaluating end separation and transverse offset losses at a graded-index (GRIN) rod lens to lens coupler is extended to the case of angular tilt misalignment loss. Two different fiber–lens combinations are examined, and the model is verified by comparison to experimental measurements, which show close agreement to the theoretical predictions. In addition, invariance of the angular tilt loss to differing optical launch and receive conditions is shown, and a possible explanation for this phenomenon is provided. Finally, the theoretical model is used to predict fiber–lens combinations which minimize possible angular tilt losses.

Wavelength division multiple access on a high-speed optical fibre LAN

Abstract A multichannel medium access control protocol suitable for implementation on a wavelength division multiplexed optical fibre LAN is discussed. Particular emphasis is placed on the mean delay against network throughput performance at individual channel transmission rates of 100 Mbit/s. The protocol is based on a token passing arbitration policy which dynamically allocates both the token and the data to individual wavelength multiplexed channels. An improved performance is demonstrated in comparison with alternative conventional approaches operating at an equivalent overall network transmission rate.

Piggybacked token-passing access protocol for multichannel optical fibre LANs

Abstract A novel Medium Access Control protocol suitable for use with multichannel optical fibre Local Area Networks employing Wavelength Division Multiplexing is described. It is based on token-passing channel arbitration in which all channels may be available to carry all of data, acknowledgement and token packets. To increase bandwidth utilization the protocol has the facility to piggyback data packets with the token at high traffic loads. Expressions are derived for the mean delay against network throughput performance for the protocol using data traffic generated with a Poisson distribution, and symmetrically disposed over the network nodes. It is demonstrated that the multichannel protocol gives a substantial improvement in network throughput performance compared with that of a conventional single channel protocol employing the same overall data transmission rate.

Performance investigation of a token passing access protocol for a multichannel optical fibre LAN

Abstract An alternative approach to multichannel local area network access is described. The new multichannel access scheme or protocol is based on a token passing arbitration policy. It is specifically intended for use with a starconfigured optical fibre LAN employing wavelength division multiplexing technology in order to provide a multichannel transmission medium. The multichannel network architecture comprises a single reservation channel and several equivalent speed data channels. Delay versus throughput performance characteristics for the network obtained by both analysis and discrete event simulation are shown. The multichannel token passing protocol is found to significantly improve these characteristics in comparison with a conventional single channel token passing approach using the same overall network transmission rate (i.e. the sum of the individual channel bit rates in the multichannel case). Furthermore, when the end-to-end propagation delay for an individual channel on the network becomes large, the multichannel token passing protocol is shown to exhibit an improved delay versus throughput performance over a multichannel protocol employing carrier sense multiple access with collision detection.

Wavelength division multiplexed optical point-sensor networks using injection laser diode sources

Abstract Wavelength division multiplexing (WDM) offers an elegant solution to the problem of increasing sensor numbers on an optical point-sensor network. In this paper the provision of WDM optical point-sensor networks is examined by determining the optical power budgets required for their implementation. A substantial increase in the number of distributed optical point-senors is demonstrated when using injection laser diode sources in conjunction with two new power-efficient WDM sensor network topologies.

Wavelength-division multiplexed multiple-sensor networks

The paper deals with some of the major issues concerned with the provision of wavelength division multiplexed multiple point-sensor networks. In particular it addresses the provision of multiple wavelength channels by the spectral slicing of an LED source. Consideration through the development of analytical models is given to slice separation, optical crosstalk, slice width, power in the slice, and, hence, the number of slices that may be obtained from a particular broadband LED source. These results are then applied to specific sensor network topologies using commercially available component parameters in order to determine the number of sensors that can be accommodated. Furthermore two novel WDM sensor network topologies are proposed which facilitate the operation of an increased number of distributed point-sensors in comparison with more conventional network topologies.

An Extrinsic Optical Fibre Sensor Employing A Partially Reflecting Mirror

A novel extrinsic optical fibre displacement sensor is described which intensity modulates an incoherent light source. Light is transmitted along multimode graded index fibre into a GRIN-rod lens coupler system. The transmission and reflection properties of this system are fully categorized. Linear movement of a partially reflecting mirror parallel to the axis of the lens coupler forms the basis of the measurement technique. This device exhibits several important features which makes it suitable for use in a process control environment. Firstly, the measurement is largely independent of the modal power distribution present at the sensor. Secondly, it can be made independent of any fluctuations in the source output power using only a single transmit and return fibre. Finally, the mechanism has a dynamic range of several millimetres which makes it eminently suitable for connection to a Bourdon tube thus providing measurement of both pressure and temperature. The characteristics of a prototype sensor in which a Bourdon tube provides conversion of the measurand to displacement of the mirror are reported, together with the requirements of the associated telemetry system. In addition a series of pressure measurements against detected optical power showing a linear relationship are presented which demonstrate the ability of the device to give an accurate measurement of both pressure and temperature.

Delay-throughput and bit-error-rate analysis for the transmission of telephony across a passive optical network using a time division multiple access/code division multiple access (TDMA/CDMA) hybrid scheme

In this paper it has been shown that by using a time division multiple access/code division multiple access (TDMA/CDMA) hybrid scheme with matched filter receivers, this will greatly improve the number of users which can be accommodated in the PON local access network architecture. This paper will also demonstrate the overall bit error rate and delay against throughput characteristics for the resulting TDMA/CDMA hybrid data stream compared with that of TDMA.

Use of extended-spectral-width broadband LEDs within WDM point-sensor networks

Spectral slicing of a broadband LED source is an efficient method for providing multiple wavelength channels for WDM sensor networking. However, optical power budgetary considerations limit the numbers that can be accommodated even with the use of power efficient topologies. Very recently, extended spectral width broadband LEDs have become available which provide spectral widths greater than 120 nm. Such LED devices exhibit both power output and spectral width capabilities comparable with the combined effect of two or three conventional LEDs. This paper reports on an investigation of the performance of such devices in relation to their utilization within WDM point-sensor networks. Furthermore, specific available device characteristics are utilized within a WDM sensor network modeling process in order to determine the maximum number of sensors that can be accommodated on various power efficient topologies. Hence, it is demonstrated that the use of these very broadband LED sources can facilitate significantly larger numbers of sensors than may be accommodated by employing spectral slicing of conventional LEDs.